JEQ Journal of Natural Resources and Life Sciences Education
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Related articles in JEQ
Right arrow Similar articles in this journal
Right arrow Similar articles in Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Web of Science (17)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Kalkhoff, S. J.
Right arrow Articles by Thurman, E. M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Kalkhoff, S. J.
Right arrow Articles by Thurman, E. M.
Agricola
Right arrow Articles by Kalkhoff, S. J.
Right arrow Articles by Thurman, E. M.
Related Collections
Right arrow Agricultural Pesticides
Right arrow Surface Water Quality
Right arrow Pesticides
Right arrow Organic Compounds
Right arrow Spatial Distribution
Journal of Environmental Quality 32:1025-1035 (2003)
© 2003 American Society of Agronomy, Crop Science Society of America, and Soil Science Society of America

TECHNICAL REPORTS
Surface Water Quality

Herbicides and Herbicide Degradation Products in Upper Midwest Agricultural Streams during August Base-Flow Conditions

Stephen J. Kalkhoff*,a, Kathy E. Leea, Stephen D. Porterb, Paul J. Terrioc and E. Michael Thurmanc

a U.S. Geological Survey, 2280 Woodale Dr., Mounds View, MN 55112
b U.S. Geological Survey, Denver Federal Center, Box 25046, MS 406, Lakewood, CO 80225-0046
c U.S. Geological Survey, 4821 Quail Crest Pl., Lawrence, KS 66049

* Corresponding author (sjkalkho{at}usgs.gov)

Received for publication February 12, 2002.

   ABSTRACT
TOP
 NOTES
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Herbicide concentrations in streams of the U.S. Midwest have been shown to decrease through the growing season due to a variety of chemical and physical factors. The occurrence of herbicide degradation products at the end of the growing season is not well known. This study was conducted to document the occurrence of commonly used herbicides and their degradation products in Illinois, Iowa, and Minnesota streams during base-flow conditions in August 1997. Atrazine, the most frequently detected herbicide (94%), was present at relatively low concentrations (median 0.17 µg L-1). Metolachlor was detected in 59% and cyanazine in 37% of the samples. Seven of nine compounds detected in more than 50% of the samples were degradation products. The total concentration of the degradation products (median of 4.4 µg L-1) was significantly greater than the total concentration of parent compounds (median of 0.26 µg L-1). Atrazine compounds were present less frequently and in significantly smaller concentrations in streams draining watersheds with soils developed on less permeable tills than in watersheds with soils developed on more permeable loess. The detection and concentration of triazine compounds was negatively correlated with antecedent rainfall (April–July). In contrast, acetanalide compounds were positively correlated with antecedant rainfall in late spring and early summer that may transport the acetanalide degradates into ground water and subsequently into nearby streams. The distribution of atrazine degradation products suggests regional differences in atrazine degradation processes.


   INTRODUCTION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
THE CORN BELT REGION of the U.S. Midwest is one of the most intensive and productive agricultural regions in the world. Nearly 80% of the USA's corn (Zea mays L.) and soybean [Glycine max (L.) Merr.] is grown in the region and more than 100000 Mg of pesticides are applied to cropland in the Midwest annually (Battaglin and Goolsby, 1999). Intensive use of agricultural chemicals may result in nonpoint-source contamination of surface and ground water throughout the Midwest. Results of studies conducted during the past decade indicate that large amounts of herbicides are washed from cropland and transported into tributary streams of major rivers of the Midwest during periods of rainfall in late spring and early summer (Thurman et al., 1991, 1992; Coupe et al., 1995; Battaglin and Goolsby, 1999). Following spring runoff of herbicides, concentrations of parent compounds in streams and rivers of the Midwest generally decrease through summer due to degradation and flushing from the soils. However, factors that affect the occurrence of herbicide compounds in small- to medium-sized streams during late summer base-flow conditions in the upper Midwest have not been thoroughly documented.

Atrazine degrades through a combination of physical, chemical, and biological processes. The three major degradation products of atrazine are deethylatrazine (DEA), deisoproplyatrazine (DIA), and hydroxy-atrazine (HA). Deisoproplyatrazine and DEA form primarily through biological processes in the soil (Kaufman and Kearney, 1970). Adams and Thurman (1991) suggest that the presence of increased DEA concentrations is primarily due to a large number of microbes, high organic carbon content, and relatively warm soil temperatures. Hydroxy-atrazine forms primarily through chemical hydrolysis (Armstrong et al., 1967) and binds to soils with high organic matter and low pH. Mesocosm studies (Lee et al., 1995) indicated that atrazine in the aquatic environment degraded more rapidly in the presence of emergent vegetation (cattail [Typha spp.]) than in open water.

Substantially less research has been conducted on the degradation of chloroacetanilide herbicides such as acetochlor, alachlor, and metolachlor. Although the application of chloroacetanilide herbicides has, at times, exceeded that of the triazine herbicides in parts of the Midwest (Meyerfeld et al., 1996), the occurrence of the chloroacetanilide herbicides has been substantially lower than triazine herbicides in surface water (Burkart and Kolpin, 1993; Thurman et al., 1992). Chloroacetanilide herbicides degrade more rapidly than atrazine (Buhler et al., 1993; Miller et al., 1997), thus reducing their potential transport to ground water or streams. Complete breakdown for all compounds, however, has not been established (Stamper et al., 1997; Miller et al., 1997). Relatively stable and persistent intermediate herbicide degradation products occur for these chloroacetanilide herbicides. Research has identified an alachlor degradation product, alachlor ethanesulfonic acid (alachlor ESA), found in both ground water and surface water (Kolpin et al., 1996, 1997; Thurman et al., 1996). Field and Thurman (1996) suggest that alachlor ESA may be the result of a glutathione conjugation process occurring in plants, algae, and terrestrial microorganisms. Additionally, it is hypothesized that mobile sulfonated and nonsulfonated degradation products of other chloroacetanilide herbicides may result from this glutathione conjugation pathway (Aga et al., 1994; Field and Thurman, 1996; Thurman et al., 1996). Outdoor mesocosm studies (Graham et al., 1999) have shown that alachlor and metolachlor were transformed to ethanasulfonic (ESA) and oxanilic acid (OA) degradation products. Alachlor OA is the more prominent breakdown product of alachlor and OA and ESA degradates of metolachlor are formed in more equal proportions in aquatic systems (Graham et al., 1999). Recent studies (Kalkhoff et al., 1998; Phillips et al., 1999) have shown that the relatively stable and soluble chloroacetanilide ESA and OA degradation products are commonly detected in streams.

Soils are an important factor in transformation and transport of triazine and chloroacetanilide herbicides in the environment. Physical and chemical characteristics of soils influence the breakdown of pesticides and their movement to ground water and nearby streams. Although deethylatrazine accounts for only a small part of atrazine degradates, it is a significant compound in surface water because of its selective removal from soil (Thurman et al., 1996). Hydroxy-atrazine is bound more tightly to the soil (Lerch et al., 1998) and may be transported to streams by soil erosion due to rainfall runoff. Fine-grained soils favor the transport of metolachlor ESA over metolachlor and metolachlor OA (Phillips et al., 1999).

Soil permeability may influence the delivery of water and pesticides to streams and affects runoff and base-flow conditions. Dissolved pesticides may reach streams through runoff or through ground water discharge into the stream depending on soil texture and slope. Pesticides may reach the stream through adsorption to sediment and subsequent transport to the stream. Areas with well-drained soils may have a greater potential for herbicide infiltration of more soluble compounds (Larson et al., 1997; Burkart et al., 1999) into ground water and subsequent transport to streams through ground water discharge. In contrast, areas with poorly drained soils have reduced infiltration to ground water and subsequently provide more runoff to streams. Poorly drained soils may be more vulnerable to erosion, however, providing more particulate forms of contaminants and sediments to enter streams. Areas with poorly drained soils in the Midwest that result from high water tables due to low relief are typically tile-drained to increase drainage. Because tile drains facilitate drainage from agricultural fields to the stream, they may short-circuit the mitigating effects of subsurface flow through riparian soils and buffer strips and serve as point sources of contaminants (Moorman et al., 1999).

Although soils may be an important factor in the transformation and transport of triazine and chloroacetanilide herbicides, few studies have documented the relation between soil characteristics in the watershed and the presence of pesticide degradates in nearby streams. Transport of herbicides to streams is of concern because these compounds may potentially affect aquatic communities (Fairchild et al., 1998, Hartgers et al., 1998) and drinking water supplies. Therefore, there is a need to better understand the persistence and transformation of heavily used herbicides the Midwest. There is also a need to understand how geological and climatological factors affect the transport of pesticides to streams and how they affect breakdown of these compounds.

The purpose of this paper is to document the occurrence of commonly used triazine and chloroacetanilide herbicides and their degradates in streams draining agricultural watersheds in southern Minnesota, eastern Iowa, and central Illinois during base-flow conditions in August 1997. The occurrence of the pesticides and pesticide degradates are related to differences in soils in the watersheds and to differences in amount of rainfall during the growing season. The results presented in this paper are one part of the U.S. Geological Survey's National Water Quality Assessment Program's (NAWQA) Midwest Regional Synoptic Study (Sorenson et al., 1999).

In Minnesota, the study area selected was the Minnesota River basin of the Upper Mississippi River basin NAWQA study unit (Stark, 1994). The study area in Iowa incorporated the entire Eastern Iowa basins NAWQA study unit (Kalkhoff, 1994). These basins included the Wapsipinicon, Cedar, Iowa, and Skunk Rivers. The study area in Illinois included the Lower Illinois River basin NAWQA study unit (Warner, 1998).


   MATERIALS AND METHODS
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Candidate sites (Fig. 1) were selected on streams at the mouth of watersheds that ranged from 150 to 2700 km2 and are located in each of three NAWQA study units (Upper Mississippi River basin, Eastern Iowa basins, and Lower Illinois River basin). In Minnesota, 24 sites were selected in watersheds located in the Minnesota River basin. Twenty-five subwatersheds were located in the Wapsipinicon, Cedar, Iowa, and Skunk Rivers in the Eastern Iowa basins. Twenty-one sites were selected in the Lower Illinois River basin. Watersheds with more than 75% agricultural land use were targeted for selection (Sorenson et al., 1999).



View larger version (40K):
[in this window]
[in a new window]
  		Fig. 1. Sampling sites and predominant soil parent materials in the study area.

 
Soils in the study area generally are moderately to poorly permeable and were formed from glacially derived material. Predominant soil parent materials, based on the USDA (1981) Major Land Resource Areas (MLRA), are shown in Fig. 1.

The most recent glaciation of Wisconsinan age deposited calcareous clay-rich tills from central and southern Minnesota into central Iowa. These deposits are commonly referred to as the Des Moines Lobe. Another glacial lobe moved into northeastern Illinois. Soil properties can be extremely variable in a watershed, but these areas contain a large proportion of poorly drained soils that were generally developed on till deposits. The Minnesota River basin and the northern part of the Eastern Iowa basins are in three MLRAs: Rolling Till Prairie (MLRA 102A), Central Iowa and Minnesota Till Prairies (MLRA 103), and the Eastern Iowa and Minnesota Till Prairie (MLRA 104). The eastern part of the Lower Illinois River basin is in the Northern Illinois and Indiana Heavy Till Plain (MLRA 110).

Older Pre-Illinoian glacial deposits have been eroded and leached for a longer period and are generally covered by a thicker layer of wind-blown loess in southern Iowa and western Illinois (Fehrenbacher et al., 1968; Anderson, 1983). Moderately permeable soils were developed on these loess deposits. The southern part of the study area in Iowa and a large part of the Illinois study area is in the Illinois and Iowa Deep Loess and Drift (MLRA 108). Areas adjacent to the Lower Illinois River and its major tributaries are in the Central Mississippi Valley Wooded Slopes (MLRA 115).

Differences between till and loess soil permeability may be mitigated to some extent by the presence of tile drains. Generally, where tile drains are present, water that has leached through the soil moves faster to nearby streams than in areas without subsurface drainage.

Rainfall was greatest in the northern part of the study area (Minnesota River basin) and decreased southeast through the Eastern Iowa basins and Lower Illinois River basin during the three months before sampling (Sorenson et al., 1999). Rainfall generally increased in the Minnesota River basin and, in contrast, decreased in the Lower Illinois River basin from May through July (Mitton et al., 1998). July rainfall was significantly less in the Eastern Iowa basins and the Lower Illinois River basin than in the Minnesota River basin, with total amounts ranging from less than 130 mm to more than 400 mm during the period of May through July (Sorenson et al., 1999). This rainfall pattern resulted in watersheds with predominantly till parent material having significantly greater rain than watersheds with predominantly loess parent material (Table 1).


View this table:
[in this window]
[in a new window]
  		Table 1. Hydrologic conditions at sites on streams draining selected agricultural watersheds in the U.S. Midwest, August 1997.

 
Stream discharge at the time of sample collection in August reflected the precipitation that fell during late spring and summer. Discharge in southern Minnesota streams ranged from near to substantially greater than the long-term average (Mitton et al., 1998) during the sampling period. In contrast, streamflow in Iowa and Illinois ranged from less than 10 to more than 50% of the long-term August monthly mean during the sampling period (May et al., 1998; Wicker et al., 1998).

Abundant rainfall and fertile soils are conducive to row-crop agriculture in the study area. Corn and soybean are the main crops (median of 75% of the watersheds is planted in corn and soybean; Sorenson et al., 1999) with corn comprising slightly greater than 50% and soybean slightly less than 50% of the cropped area. Although the amount of each watershed planted in corn was not significantly different, the subbasins sampled in the Minnesota River basin had a small but statistically significantly greater amount of soybean acres than in the sampled Eastern Iowa and Lower Illinois River subbasins.

Herbicides used to control competing vegetation in corn and soybean vary by state and may vary by soil parent material. Atrazine and cyanazine were used on substantially less crop area and at a substantially lower rate in Minnesota than in Iowa and Illinois in 1997 (Table 2). Metolachlor is used on more crop area in Iowa than in Illinois and Minnesota. Data shown in Table 2 are state-wide averages and probably are the best available data, but do not necessarily represent the usage rates in each watershed. Several reports suggest that herbicides are not uniformly applied within each state. Fallon et al. (1997) illustrated that atrazine use (on a county level) increased from northwest to southeast in the Minnesota River basin. Stoltenberg and Pope (1990) reported that atrazine was applied to only 25% of the corn in north-central Iowa in contrast to 69% of the corn in southeastern Iowa during the 1980s. Application rates also were less in north-central Iowa in areas with lower-permeability till soils that contained greater organic carbon content because of farmers' concern that "carryover" of atrazine residues in soils may damage soybean. Although data are not available, the concern for atrazine carryover also may result in reduced application rates by farmers in areas of Minnesota and Illinois with till soils. Atrazine use data reported in Minnesota (Table 2) appear to reflect this pattern. Also, in response to the detection of atrazine in ground water, atrazine management areas were established in northeastern Iowa in areas where bedrock aquifers are close to landsurface (Iowa Department of Agriculture and Land Stewardship, 1999). No more than 1.7 kg of atrazine per hectare may be applied per year in these management areas. Pesticide use data suggest that farmers alternatively applied alachlor, metolachlor, and most recently acetochlor to control competing vegetation in pesticide-management areas.


View this table:
[in this window]
[in a new window]
  		Table 2. Use of selected triazine and chloroacetanilide herbicides on corn and soybean in Iowa, Illinois, and Minnesota, 1997.{dagger}

 
Sample Collection
Sampling for analysis of pesticides and pesticide degradates took place concurrently in the three National Water-Quality Assessment Program study units during August 1997. Sampling methods are described in detail by Sorenson et al. (1999) and are briefly summarized here. Water was collected with a depth-integrated sampler at 5 to 10 verticals equally spaced across the stream or river (Edwards and Glysson, 1988; Ward and Harr, 1990; Shelton, 1994) to integrate vertical and horizontal water quality variability. Samples were then filtered through baked glass-fiber filters with a nominal 0.7-µm pore diameter to remove suspended particulate matter (Shelton, 1994). Following field processing, all samples were immediately chilled (placed in a cooler with wet ice) for shipment to the analytical laboratories. Sampling and filtering equipment were decontaminated between samples by washing in a detergent solution, rinsing with deionized water, and finally rinsing with methanol. After air-drying, the equipment was stored in plastic bags or wrapped in aluminum foil in preparation for transport to the next sampling site (Shelton, 1994).

Laboratory Analysis
The U.S. Geological Survey's Organic Geochemistry Research Laboratory (OGRL) in Lawrence, Kansas, analyzed water samples. The parent compounds (ametryn, atrazine, alachlor, acetochlor, cyanzine, metolachlor, metribuzin, prometon, prometryn, propachlor, propazine, simazine, and terbutryn) and the triazine degradation compounds (cyanazine-amide, deethylatrazine, and deisoproplyatrazine) (Table 2) were analyzed with gas chromatography–mass spectrometry following extraction on C18 cartridges (Meyer et al., 1993; Thurman et al., 1990; Zimmerman and Thurman, 1999). The analytical reporting limit for this method was 0.05 µg L-1 for all compounds. Six chloroacetanilide herbicide degradation products (acetochlor ESA, acetochlor OA, alachlor ESA, alachlor OA, metolachlor ESA, and metolachlor OA) and the atrazine degradate (hydroxy-atrazine) (Table 3) were analyzed by high-performance liquid chromatography (HPLC) following solid-phase extraction on C18 cartridges (Ferrer et al., 1998; Zimmerman et al., 2000). Approximately 5% of the samples were verified with HPLC–MS using the method of Ferrer et al. (1998). This combination of methods was able to clearly distinguish between alachlor ESA and acetochlor ESA in the samples. The analytical reporting limit for this method was 0.2 µg L-1.


View this table:
[in this window]
[in a new window]
  		Table 3. Herbicides and herbicide degradation products analyzed.

 
Although the analytical reporting limits for the parent compounds and three atrazine degradation products were four times lower (0.05 versus 0.20 µg L-1) than the acetanalide degradation products under investigation, the uncensored data were used when comparing the frequencies of detection among compounds. Loss of a substantial amount of data would occur if a common detection threshold of 0.20 µg L-1 were used to compute the frequency of detection for a compound. The result may be an underestimation of the frequency of detection of the acetanalide degradation products in relation to the parent compounds. Although not known, lower analytical detection limits probably would have increased the frequency of detection for the degradate compounds. Also, because frequency of detections of the degradation products generally were greater than the parent compounds (see Results), the patterns of detection would not have been changed with lower frequency of detections but probably would have been strengthened.

Data Analysis
Data from all 70 sites were used in summarizing the occurrence of the herbicide compounds. Because of the nonnormal distribution of the data, comparison between groups of data was made with the nonparametric Kruskal–Wallis test and correlations between constituents were made with the Spearman's Rho test (Helsel and Hirsch, 1992).


   RESULTS
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Because of their chemical and physical properties, commonly used triazine and chloroacetanilide herbicides were expected to degrade by biological and chemical processes through the growing season. Degradation should be evident by the presence (or absence) of herbicide compounds in rivers and streams that drain agricultural areas of the Midwest. Degradation should also be reflected in the amount of the parent compound in relation to degradation compounds. The concentration of degradate compounds in streams may be influenced by conditions in the watershed, such as soil properties and rainfall, that affect breakdown and transport. Dilution by ground water free of pesticides also will influence degradate concentrations.

During the study, both triazine and chloracetanalide residues (parent plus degradation products) were detected at the 70 sites in the upper Midwest in August 1997 (Table 4). Among all sites, 6 parent compounds and 10 degradation products were detected. Ametryn, alachlor, metribuzin, prometryn, propachlor, propazine, and terbutryn were not detected at the 0.05 µg L-1 level and, with the exception of alachlor, are not discussed in the remainder of this report. The frequency of detection of the parent compounds was similar to use: atrazine > metolachlor > cyanazine. Atrazine, metolachlor, and cyanazine were the most frequently detected herbicides and were present in 94, 59, and 37% of the samples, respectively. Although not listed among the most heavily used herbicides (USDA, 1998), prometon and simazine were detected in 16 and 7% of the samples, respectively. Prometon is generally used to control vegetation in noncrop areas that include paved roads, rights-of-way, and industrial sites that may be susceptible to erosion. Although its total use is probably substantially less than other herbicides in the Midwest, prometon is much more persistent (aerobic soil half-life is 932 d) and is commonly applied at substantially greater rates (Capel et al., 1999). Simazine is used for agricultural purposes, but it also commonly occurs in samples with prometon (Gilliom et al., 1999), suggesting an urban or nonagricultural source.


View this table:
[in this window]
[in a new window]
  		Table 4. Statistical summary of herbicide compounds in 70 agricultural streams of the U.S. Midwest, August 1997.

 
Seven of nine compounds detected in greater than 50% of the samples were herbicide degradation products (Table 4). Metolachlor degradation products were two of the four most commonly detected degradation products. Metolachlor ESA was detected in 96% of the samples and was detected more often than any other compound. Metolachlor OA was detected in 80% of the samples. Surprisingly, alachlor ESA, a degradate of a herbicide that has low use and that was not detected in the streams, was the second-most commonly detected compound. The frequent presence of alachlor ESA in these streams may be the legacy of the heavy historical use of alachlor (Meyerfeld et al., 1996). The four triazine degradation products were detected in 41 to 81% of the samples (Table 4).

Concentrations of parent compounds were generally low, ranging from less than 0.05 to 1.5 µg L-1, and did not exceed established drinking water standards (USEPA, 2000). Several pesticides analyzed do not have established standards. With two exceptions, atrazine and prometon, the parent compounds were present at concentrations less than 1.0 µg L-1. Median concentrations were less than the analytical reporting limit for all parent compounds except atrazine (0.17 µg L-1) and metolochlor (0.06 µg L-1).

Degradate concentrations ranged from less than 0.2 to 8.8 µg L-1. The ethanesulfonic acid degradation products were generally present in greater concentration than the triazine degradation products. In addition to being detected most frequently, metolachlor ESA was present in the greatest concentration (median of 1.7 µg L-1). With the exception of alachlor ESA (0.58 µg L-1), median concentrations of the remaining compounds were less than 0.35 µg L-1.

Metolachlor and its degradation products were the prevalent pesticide compounds (by mass) in streams in the Midwest in August 1997. On average, greater than 50% of the total pesticide residue (by mass) consisted of metolachlor compounds (parent and degradation products). Alachlor and atrazine compounds each contributed about 14%, acetochlor about 9%, and cyanazine compounds less than 1% of the total pesticide mass.

Although chemical properties such as solubility and degradation rates are factors, the greater metolachlor compound mass can also be partially attributed to the high rate of metolachlor usage (Table 2). In addition to being frequently detected, alachlor degradation products were present in surprisingly high concentrations based on the amount of parent compound applied in 1997. The disproportionately high concentration in relation to use suggests that alachlor degradation products are persistent and may originate from applications in previous years. Kolpin et al. (1998) have shown that alachlor ESA is one of the most commonly detected pesticide degradation products in shallow ground water.


   DISCUSSION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
CONCLUSIONS
 REFERENCES
 
Relation between Parent Compounds and Degradation Products
In August 1997, after pesticide application, degradation products were the most frequently detected herbicide compounds and were present in significantly greater concentrations than the parent compounds in streams of the Midwest. The median value for the summed concentrations of the degradation products in each sample (4.4 µg L-1) was almost 17 times greater than the median value for the summed concentrations of the parent pesticides (0.26 µg L-1). During the study, multiple herbicide compounds were commonly detected in each stream. Multiple parent herbicide compounds were detected in 70% of the samples and multiple degradate compounds were detected in all samples. The median number of parent pesticide compounds detected in the samples was two and the median number of pesticide degradation products detected in the samples was seven.

The composition of the chloroacetanilide residue was substantially different than the composition of the triazine residues in August 1997. Chloroacetanilide compounds in the streams consisted almost exclusively of degradation products. The total acetochlor residue (parent plus degradation products) contained less than 1% parent compound. Alachlor residue consisted entirely of the ESA and OA degradation products. Metolachlor residue consisted of about 2% parent and about 98% degradation products. In contrast, the triazine compounds are present as a mixture (25–50%) of parent and degradation products. Both atrazine and cyanazine degrade biologically to DIA. However, it is assumed for this study that the DIA originates from atrazine due to the difficulty in partitioning the source and to the fact that atrazine use was substantially greater than cyanazine. Because of this assumption, atrazine degradation products may be slightly overestimated and cyanazine degradation products slightly underestimated. Cyanazine residue consisted of 50% parent and 50% cyanazine-amide. Reddy et al. (1997) found that cyanazine-amide was more likely to remain in the aqueous phase and thus have greater transport potential by water than cyanazine. Atrazine residue consisted of about 25% parent, 25% DIA plus DEA, and about 50% HA.

Although HA was commonly present in streams throughout the Midwest, it was more prevalent in the southern part of the study area. Generally, concentrations of HA would be expected to be higher in streams draining basins with higher atrazine application rates, but increased concentrations also may result from differing degradation pathways. The ratios of degradation compound to parent compound were calculated to compensate for the differing application rates (Fig. 2) . If atrazine degraded to HA, deethylatrazine, and deisoproprylatrazine uniformly, the degradate to the parent compound ratios would be uniform across the study area. These ratios were not uniform. The hydroxy-atrazine to atrazine ratio (HAR) increased and the deethylatrazine plus deisoproprylatrazine to atrazine ratio (DDAR) decreased from north to south (Fig. 2). An increasing HAR with a concurrent decreasing DDAR from southern Minnesota to central Illinois suggests atrazine degradation is not consistent across the Midwest.



View larger version (23K):
[in this window]
[in a new window]
  		Fig. 2. Relation between hydroxy-atrazine to atrazine ratio (HAR) and deethylatrazine plus deisoproplyatrazine to atrazine ratio (DDAR) and the latitude.

 
Relation between Herbicide Compounds and Soil Parent Material
Pesticide compounds found in streams of the Midwest during late summer 3 to 4 mo after application were partially dependent on the predominant soil parent material in the watershed. The two dominant classes of herbicides (triazines and acetanalides) considered in this study were detected at different rates and at different concentrations, dependent on whether till- or loess-type soils predominated in the watershed. Chemical properties of these herbicides may influence their use and subsequent detection in basins with different soil types.

Atrazine and cyanazine, the two most heavily used triazine herbicides in the Midwest, and several of their primary degradation products were present more frequently in streams draining watersheds with predominantly loess-type soils than in watersheds with predominantly till soils (Fig. 3) . Although the frequency of detection for atrazine was not significantly different, atrazine degradation products were detected more frequently in streams draining loess soils than in streams draining till soils.



View larger version (32K):
[in this window]
[in a new window]
  		Fig. 3. Frequency of detection and median concentration of selected herbicides and herbicide metabolites in streams draining till and windblown loess-derived soils of the upper U.S. Midwest, August 1997.

 
Field and laboratory studies suggest that soil composition and temperature may influence which compound is produced through atrazine degradation. Early laboratory research (Hance and Segal, 1980) found that greater concentrations of HA were formed in acidic rather than neutral soils. Although HA leached from soils in small amounts in relation to atrazine, Mersie and Seybold (1996) found that almost twice as much HA desorbed from a low clay content, lower pH, and lower organic carbon content soil similar to loess soils. Hydroxy-atrazine adsorbs readily to soil and may be transported to the streambed during summer runoff. Desorption by ground water seeping into streams through contaminated streambed sediment has been postulated as a source of HA in streams during base-flow conditions (Lerch et al., 1998).

In contrast, two ESA degradation products, alachlor ESA and acetochlor ESA, were detected more frequently in samples from streams draining till soils than in streams draining loess soils (Fig. 3). Several chloroacetanilide degradation products (alachlor ESA, metolachlor ESA, metolachlor OA, and acetochlor ESA) were present in significantly (P < 0.05) greater concentrations in samples from streams draining watersheds with till soils than watersheds with loess soils.

Herbicidal properties, such as degradation rate, may determine application rates, which in turn partially account for the presence of herbicides in streams during late-summer base-flow conditions. Triazine pesticide use rates are apparently less in areas with predominantly till soils (Stoltenberg and Pope, 1990) than in areas of predominantly loess soils. This is due to the relatively high pH and organic carbon content of till soils that hinder the degradation of triazine herbicides, resulting in the "carryover" of atrazine in the soil to the following growing season that may harm soybeans in a corn–soybean crop rotation. Because of lower use on till soils, triazine degradation products would be expected to be lower. Cyanazine sorption has been correlated with fine soil texture and greater organic carbon content (Reddy et al., 1997). Alachlor, metolachlor, and acetochlor possibly were used to offset triazine pesticide reductions in areas with till soils. Acetanalide herbicides also are adsorbed to organic matter (Miller et al., 1997), but because of their generally shorter half-life (Barbash et al., 1999), they are not present in substantial quantities in soils the following growing season.

Relation between Herbicide Compounds and Rainfall
Triazine and chloroacetanilide herbicide transport to streams of the Midwest in late summer was influenced by the timing and amount of rainfall during the growing season. Concentrations of metolachlor compounds in streams (consisting primarily of degradation products) were greatest in streams that received greater amounts of rainfall early in the growing season (Fig. 4) . In contrast, concentrations of atrazine compounds (consisting of a mixture of parent and degradation products) were least in streams draining watersheds that received the most rainfall during the late summer (Fig. 4). The sum of the concentrations of atrazine compounds commonly exceeded 2 µg L-1 in streams draining watersheds that received less than 100 mm of rain in June and July, and rarely exceeded 1 µg L-1 in streams draining watersheds that received more than 200 mm of rain in June and July.



View larger version (24K):
[in this window]
[in a new window]
  		Fig. 4. Concentrations of atrazine and metolachlor compounds in relation to rainfall during selected periods in the watershed during the growing season.

 
Although atrazine and metolachlor adsorb to soil materials in similar amounts (Mersie and Seybold, 1996), metolachlor has been found to be less susceptible to runoff transport than atrazine (Jaynes et al., 1999). Initially, chloroacetanilide pesticides adsorb to soil and may not be readily available to transport to streams and leach to ground water. Chloroacetanilide pesticides probably then degrade in the soils within weeks of application. As the pesticides degrade to the more soluble ESA and OA degradation products, summer rains may transport degradation products to streams and ground water. Kolpin et al. (1998) have documented the transport of metolachlor and alachlor degradation products to shallow ground water. As a result, degradation products will be present in ground water and potentially available for transport to streams during late summer.

Since the triazine pesticides are generally more soluble than the chloroacetanilide pesticides, excess rainfall may have flushed much of the atrazine from the soil early in the growing season. Atrazine concentrations have been found to be greatest in small (Kalkhoff and Schaap, 1996) and large (Clark et al., 1999) streams and rivers of the Midwest during May and June when rainfall runoff is greatest. However, atrazine is persistent (half-life of more than 140 d; Barbash et al., 1999) and any remaining after the spring flush, for the most part, has not degraded and is available to be transported from the soil. Atrazine may be flushed both into nearby streams and transported to ground water because of its relatively high solubility.

Although water originating from rainfall may require several decades to move through local sand and gravel aquifers to streams (Puckett and Cowdery, 2002), part of the ground water contributing to flow in the streams studied during this investigation was in aquifers for only a short period of time. The age of ground water entering these 70 streams is difficult to accurately quantify without detailed hydrologic studies, but may be a mix of old water contributed from deeper regional flow and younger water from shallower localized flow systems. Two lines of evidence suggest that at least part of the water in the streams of the Midwest in August is relatively young ground water. The presence of acetochlor or acetochlor degradates in 75% of the streams indicates that at least part of the water was three years old or younger. Acetochlor was registered for use beginning in the 1994 growing season, three years before this investigation began. The correlation of metolachlor and metolachlor degradates with late spring and early summer rains (Fig. 4) also suggests that ground water originating as rainfall during the current growing season was part of the streamflow in the streams of the Midwest. This correlation would not be expected if water transporting pesticides applied during the 1997 growing season were not present in the streams.

The young ground water that comprises part of the streamflow may originate from water infiltrating soil relatively close to the stream (Puckett and Cowdery, 2002) that then moves through the shallow water table to the streams. Crops are commonly grown on the flat flood plain and pesticides are applied in close proximity to many of the streams of the Midwest. In addition, young ground water may also originate from considerable distance from the stream, being transported to the stream in tile lines. Phillips et al. (1999) have shown that pesticides and pesticide degradates are rapidly transported from the soil to the water table after rain events. Since tile lines are generally located in or near the top of the water table, recently recharged water containing any dissolved pesticide compounds will be routed to a nearby stream.


   CONCLUSIONS
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
REFERENCES
 
Though the Corn Belt region of the U.S. Midwest is commonly thought of as a relatively homogeneous area (Omernick, 2000), relatively small differences (in relation to the rest of the United States) in soils and climate have a substantial influence on the amount and types of herbicide compounds that are transported to streams that drain corn- and soybean-producing areas. By late summer of 1997, several months after application, the majority of the herbicide residues in streams of the Midwest are degradation compounds. Degradation products of the commonly used triazine (atrazine and cyanazine) and chloroacetanilide (acetochlor, alachlor, metolachlor) herbicides comprise the majority of pesticide compounds found in the streams during base-flow conditions in late summer. Although degradation products constitute a major part of the residue, the two classes of herbicides have substantially different compositions. The chloroacetanilide residue was comprised almost exclusively of degradate compounds, but the triazine residue consisted of a mixture of parent and degradate compounds.

As would be expected, differences in use patterns contributed to differences in the types of pesticides present in late summer in streams of the Midwest. Atrazine and to some extent cyanazine compounds are more prevalent in watersheds with a majority of moderately permeable loess soils than in watersheds with mostly poorly permeable till soils. The predominant atrazine degradates change from the biologically derived DIA and DEA in the north to the chemically and biologically derived hydroxy-atrazine in the southern part of the Midwest study area.

The timing of rainfall during the growing season is an important factor in transport of pesticide residue to streams during late summer. Increased rainfall early in the growing season (June) contributed to increased concentrations of metolachlor degradation products in streams during August, whereas increased rainfall during mid-summer (July) flushed most atrazine compounds from the watersheds, resulting in lower concentrations in August.

Results of this study indicate that even though concentrations of commonly used triazine and chloroacetanilide herbicides are low in streams of the Midwest during late summer, some of their degradation products are present in substantial concentrations. Initial studies (Heydens et al., 1996) suggested that ESA degradation products are less toxic than the parent compounds. However, studies documenting toxicological effects of the OA degradation products and effects of chronic exposure on aquatic organisms are sparse.

Scarcity of detailed herbicide use information (county or watershed level) complicates the interpretation of water quality results from similar geologic areas in the Midwest. Most of the degradation products studied were the initial breakdown products and the presence and concentration of degradation products further down the breakdown pathway in streams is unknown. Additional studies to document and quantify these secondary and tertiary degradation products are needed to fully understand the fate and movement of pesticides through the hydrologic system.


   ACKNOWLEDGMENTS
 
The authors would like to thank the U.S. Geological Survey's National Water-Quality Assessment Program National Leadership Team for providing additional funds for the Midwest Regional Synoptic Study, and Steve Sorensen and study unit biologists Mitch Harris and Linda Roberts, who were able to persevere in the face of many setbacks and obstacles to make the synoptic study a reality. Appreciation is expressed to the numerous USGS field personnel in Illinois, Iowa, and Minnesota who spent many hours in the streams collecting samples and gathering information, without which this report would have not been possible. T. Steinheimer of the USDA-ARS Soil Tilth Laboratory and S. Larson and D. Dupre of the U.S. Geological Survey reviewed earlier drafts of the paper. Suggestions by three anonymous reviewers greatly improved the report.


   NOTES
TOP
 NOTES
ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Abbreviations: DEA, deethylatrazine; DIA, deisoproplyatrazine; ESA, ethanesulfonic acid; HA, hydroxy-atrazine; OA, oxanilic acid.


   REFERENCES
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 


Related articles in JEQ:

This Issue in Journal of Environmental Quality

JEQ 2003 32: 745-750. [Full Text]  



This article has been cited by other articles:


Home page
 Journal of Soil and Water ConservationHome page
E.A. Pappas, C. Huang, and D.L. Bucholtz
Implications of sampling frequency to herbicide conservation effects assessment
Journal of Soil and Water Conservation, November 1, 2008; 63(6): 410 - 419.
[Abstract] [PDF]

Home page
 J. Environ. Qual.Home page
M. S. Dalton and E. A. Frick
Fate and Transport of Pesticides in the Ground Water Systems of Southwest Georgia, 1993-2005
J. Environ. Qual., September 2, 2008; 37(5_Supplement): S-264 - S-272.
[Abstract] [Full Text] [PDF]

Home page
 J. Environ. Qual.Home page
L. J. Krutz, C. H. Koger III, M. A. Locke, and R. W. Steinriede Jr.
Reduced Surface Runoff Losses of Metolachlor in Narrow-Row Compared to Wide-Row Soybean
J. Environ. Qual., July 17, 2007; 36(5): 1331 - 1337.
[Abstract] [Full Text] [PDF]

Home page
 J. Environ. Qual.Home page
M. Larose, G. C. Heathman, L. D. Norton, and B. Engel
Hydrologic and Atrazine Simulation of the Cedar Creek Watershed Using the SWAT Model
J. Environ. Qual., March 1, 2007; 36(2): 521 - 531.
[Abstract] [Full Text] [PDF]

Home page
 J. Environ. Qual.Home page
L. J. Puckett and W. B. Hughes
Transport and Fate of Nitrate and Pesticides: Hydrogeology and Riparian Zone Processes
J. Environ. Qual., November 7, 2005; 34(6): 2278 - 2292.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Related articles in JEQ
Right arrow Similar articles in this journal
Right arrow Similar articles in Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Web of Science (17)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Kalkhoff, S. J.
Right arrow Articles by Thurman, E. M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Kalkhoff, S. J.
Right arrow Articles by Thurman, E. M.
Agricola
Right arrow Articles by Kalkhoff, S. J.
Right arrow Articles by Thurman, E. M.
Related Collections
Right arrow Agricultural Pesticides
Right arrow Surface Water Quality
Right arrow Pesticides
Right arrow Organic Compounds
Right arrow Spatial Distribution

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
The SCI Journals Agronomy Journal Crop Science
Journal of Natural Resources
and Life Sciences Education		 Vadose Zone Journal
Soil Science Society of America Journal Journal of Plant Registrations The Plant Genome